Facial symmetry is the single most repeated claim in online beauty content. Flip a photo of your face, mirror each half, and you'll find a hundred articles explaining that the more identical those halves are, the more attractive you are.
The claim isn't wrong. It's just wildly oversold — and the way it's usually presented causes real distress to people whose faces are, like every human face ever measured, slightly asymmetrical.
Here's what four decades of research actually establishes, what it doesn't, and how to think about your own result if you've taken a face symmetry test.
The claim, stated precisely
The scientific version of the symmetry hypothesis is narrower than the internet version.
Researchers distinguish between two kinds of asymmetry. Directional asymmetry is systematic — a population-level tendency for one side to differ from the other in a consistent direction. Human hearts are on the left; that's directional. Fluctuating asymmetry is random deviation from perfect bilateral symmetry, differing unpredictably in direction and magnitude between individuals.
Only fluctuating asymmetry is theoretically interesting for attractiveness. The reasoning comes from developmental biology: a bilaterally symmetrical body plan is genetically specified, so random deviations from it may indicate that development was perturbed — by pathogens, malnutrition, toxins, or genetic load. If fluctuating asymmetry is a readable trace of developmental stress, then a preference for symmetry could be an evolved preference for developmental stability.
That's the hypothesis. It's elegant, and it generated an enormous research programme.
What the evidence supports
The core finding replicates. In her review in the Annual Review of Psychology, Gillian Rhodes conducted critical review and meta-analyses of the candidate biologically-based beauty standards and found that averageness, symmetry, and sexual dimorphism are all attractive in male and female faces — and across cultures.
That cross-cultural component matters. Rhodes and colleagues had earlier tested averageness and symmetry preferences in non-Western populations specifically to check whether these were Western media artefacts. They weren't. Preferences appeared in cultures with minimal exposure to Western beauty imagery, which is difficult to explain if the preference were purely learned convention.
Developmental evidence points the same way. Preferences for attractive faces emerge remarkably early — infants show looking-time preferences for faces adults rate as attractive, well before any plausible cultural conditioning. Rhodes notes that this early emergence, combined with cross-cultural agreement, challenges the long-standing view that attractiveness preferences simply reflect arbitrary cultural standards.
Symmetry also shows up in the experimental manipulation studies. When researchers take a single face and produce artificially more and less symmetrical versions, holding everything else constant, raters reliably prefer the more symmetrical version. Perfectly symmetric versions created by blending a face with its mirror image are preferred over less symmetric versions of the same face — including when both versions are blends, which controls for the smoothing artefacts that blending introduces.
So: symmetry preference is real, replicated, cross-cultural, present early, and survives careful experimental controls.
What the evidence does not support
Now the corrections — and there are several important ones.
The effect size is modest
"Statistically reliable" and "practically dominant" are different things. Symmetry is one predictor among several, and in most studies it explains a modest share of the variance in attractiveness ratings. Averageness typically performs at least as strongly, and there's a good reason for that: averaging faces also makes them more symmetrical, which is why researchers had to work hard to disentangle the two effects.
If you rank the drivers of facial attractiveness by how much rating variance they account for, symmetry is on the list — but it is not at the top, and it is nowhere near the whole story.
Normal asymmetry is largely invisible
This is the correction that matters most to actual people.
The asymmetries used in experiments are often larger than those found in ordinary faces. Researchers manipulate faces to create clearly distinguishable symmetric and asymmetric versions precisely so that raters can perceive the difference. That's methodologically necessary — and it means the experimental effect doesn't straightforwardly transfer to the millimetre-scale variation present in typical faces.
There is reason to think observers simply aren't tuned to detect subtle asymmetry. The visual system is extremely good at holistic face processing and comparatively poor at precise metric comparison between the two halves of a face seen simultaneously.
Practically: pronounced asymmetry — a significantly deviated nasal septum, marked facial nerve weakness, one eyelid noticeably lower — is perceptible and does influence perception. The 2mm difference in eye height that a symmetry tool reports and you have now noticed for the first time is not.
The mirror test is misleading
The standard demonstration — mirroring each half of your face and comparing the two resulting composites — produces images that look strange and often unappealing. People take this as proof of significant asymmetry.
It isn't. Two things are happening. First, mirroring doubles any asymmetry: a 2mm difference becomes a 4mm difference between the two composites. Second, and more importantly, the composites are unfamiliar images, and unfamiliarity itself reads as wrongness in face perception. Your brain has a highly tuned representation of what your face looks like. Any deviation from it — including a perfectly symmetrical version — triggers a mismatch signal you experience as "that looks off."
Every mirrored composite of every human face looks strange. This tells you about face perception, not about your appearance.
The health signal is contested
The evolutionary rationale — that symmetry signals developmental stability and therefore health — is the weakest link in the chain.
Research testing whether facial symmetry actually predicts health outcomes has produced mixed results. Some studies find associations between facial symmetry, averageness, and measures of health or perceived health; others find null or weak effects. Rhodes and colleagues examined whether averageness and symmetry genuinely signal health and found the relationship less robust than the theory requires.
The current state of play is roughly: people behave as though symmetry signals quality, and the perceptual preference is well documented, but whether it reliably tracks actual health is unresolved. It's entirely possible that the preference is partly a by-product of how visual systems process symmetric stimuli generally — symmetry preference isn't unique to faces, and appears in aesthetic judgments of many object categories.
Why your face is asymmetrical (and everyone else's is too)
Perfect bilateral symmetry does not occur in humans. Several ordinary processes guarantee this:
Developmental noise. Cell division and tissue growth are stochastic processes. Two sides developing under identical genetic instruction still diverge slightly.
Chewing side preference. Most people chew predominantly on one side, producing modest differences in masseter muscle bulk over decades.
Sleeping position. Consistent side-sleeping applies sustained pressure to one side of the face, with measurable soft-tissue effects over years.
Expression habits. A characteristic one-sided smile or a habitual raised eyebrow shapes the underlying musculature asymmetrically over time.
Dental history. Extractions, orthodontics, and unilateral tooth loss all alter facial support.
Normal skeletal variation. The two halves of the facial skeleton simply differ slightly in almost everyone.
Faces you consider beautiful are asymmetrical. Faces on magazine covers are asymmetrical — often noticeably so once you look for it. Several widely admired actors have pronounced facial asymmetries that are part of what makes them distinctive rather than a defect.
How symmetry is actually measured
If you want to know your own symmetry index rather than guess from a mirror trick, the measurement is straightforward in principle.
A landmark-based system identifies corresponding points on each side — outer and inner eye corners, alar bases of the nose, mouth corners, jaw angle points, brow positions — then establishes the facial midline, typically as a best-fit vertical axis through central landmarks like the nasion and the midpoint of the philtrum.
For each pair of bilateral landmarks, the system computes the deviation between the point's position and its mirror-image position across that midline. These deviations are aggregated — usually normalised by face size so that a large face and a small face are comparable — into a single index.
Two practical caveats:
Head rotation destroys the measurement. A face rotated even a few degrees from frontal produces perspective differences between the two sides that read as asymmetry. Any symmetry measurement from a non-frontal photo is largely measuring your head angle.
Camera distance matters too. As documented in JAMA Facial Plastic Surgery, short-distance photography meaningfully distorts facial proportions — a 12-inch selfie makes the nasal base appear roughly 30% wider relative to the face than the same face shot from five feet. Distortion isn't necessarily symmetrical if your head is even slightly turned.
For a meaningful symmetry reading: frontal, neutral expression, even lighting, camera at least three feet away, eyes level with the lens.
What to do with your symmetry score
If your score is high: it's one favourable structural trait among several. It isn't a licence to ignore the others, and it isn't a guarantee of anything.
If your score is middling: you are where nearly everyone is. Middling symmetry is the human default, and the deviation is almost certainly below what observers perceive.
If your score is low: first, retake the measurement with proper frontal alignment, because head rotation is the most common cause of an artificially low reading. If it remains low across properly-taken photos, you have pronounced asymmetry — which is worth knowing, occasionally worth investigating medically (sudden or progressive asymmetry warrants a doctor's opinion), and still not a verdict on your appearance.
The one thing not to do is start monitoring it. Repeatedly measuring and comparing facial features is a documented pattern in appearance-focused online communities, and it correlates with worse rather than better outcomes for how people feel about their faces.
The honest summary
Facial symmetry has a genuine, replicated, cross-culturally consistent effect on attractiveness ratings. It is one of the better-supported findings in the field.
It is also a modest effect, largely operative at levels of asymmetry beyond what most faces exhibit, and considerably less important than the internet's treatment of it suggests. The evolutionary story behind it remains contested. And the mirror test that convinced you your face is lopsided is a demonstration of face-perception quirks, not a measurement.
If you want an actual measurement rather than a mirror trick, take our face symmetry test. If you want the broader picture of what drives facial attractiveness, start with what makes a face attractive — symmetry is on the list, but it shares the list with several traits that matter at least as much.
Frequently Asked Questions
Is anyone's face perfectly symmetrical?
No. Developmental noise, chewing side preference, sleeping position, expression habits, and normal skeletal variation guarantee some asymmetry in every human face. Perfect bilateral symmetry does not occur naturally.
Does facial symmetry really affect attractiveness?
Yes, but modestly. Meta-analytic reviews confirm symmetry is attractive in both male and female faces and across cultures. However, the effect is one of several predictors, and the asymmetries used in experiments are often larger than those found in ordinary faces.
Why does my mirrored face look so strange?
Two reasons. Mirroring doubles any existing asymmetry, and the resulting image is unfamiliar — your brain has a precise stored representation of your actual face, so any deviation reads as wrong. Every person's mirrored composite looks strange.
Can I make my face more symmetrical?
Soft-tissue asymmetry can shift modestly with habit changes such as sleeping position and chewing side, though effects are small and slow. Skeletal asymmetry is fixed without surgery. Given the modest effect of typical asymmetry on perception, this is rarely worth pursuing.
Should I worry about a low symmetry score?
Usually not — first check whether your photo was properly frontal, since head rotation is the most common cause of an artificially low score. Sudden or progressive facial asymmetry is a medical question worth raising with a doctor; longstanding mild asymmetry is normal human variation.
References
- Rhodes, G. (2006). The Evolutionary Psychology of Facial Beauty. Annual Review of Psychology, 57, 199–226. https://doi.org/10.1146/annurev.psych.57.102904.190208
- Rhodes, G., Yoshikawa, S., Clark, A., Lee, K., McKay, R., & Akamatsu, S. (2001). Attractiveness of facial averageness and symmetry in non-Western cultures. Perception, 30(5), 611–625. https://doi.org/10.1068/p3123
- Langlois, J. H., Kalakanis, L., Rubenstein, A. J., et al. (2000). Maxims or Myths of Beauty? A Meta-Analytic and Theoretical Review. Psychological Bulletin, 126(3), 390–423. https://doi.org/10.1037/0033-2909.126.3.390
- Little, A. C., Jones, B. C., & DeBruine, L. M. (2011). Facial attractiveness: evolutionary based research. Philosophical Transactions of the Royal Society B, 366(1571), 1638–1659. https://doi.org/10.1098/rstb.2010.0404
- Ward, B., Ward, M., Fried, O., & Paskhover, B. (2018). Nasal Distortion in Short-Distance Photographs: The Selfie Effect. JAMA Facial Plastic Surgery, 20(4), 333–335. https://doi.org/10.1001/jamafacial.2018.0009